An integrative analysis of ctDNA fragmentomics and methylation across breast cancer subtypes to identify subtype-specific prognostic biomarkers

Abstract

In cancer patients, a substantial fraction of cell-free DNA (cfDNA), derived from tumor cells, offers a non-invasive opportunity to interrogate tumor biology. This study aimed to extract, profile and identify unique fragmentation and methylation patterns in four different breast cancer subtypes (HER2+, Luminal B, Triple Negative Breast Cancer, Triple Positive Breast Cancer) compared to healthy controls. Furthermore, it utilizes the fragmentation feature profiles along with epigenetic analysis to evaluate their potential for subtype -specific detection. Public cfDNA sequencing datasets from healthy controls and four types of breast cancer subtypes cases were retrieved from the NCBI SRA repository. Fragmentation features including Fragment Size Distribution (FSD), Fragmentation Size Ratio (FSR), Fragmentation Size Coverage (FSC), BreakPoint Motif (BPM), End Motif (EDM), Copy Number Variation (CNV) and epigenetics profiles like methylation profiles were searched using a python based workflow and downstream analyzed in R. Our findings indicated subtype specific distinctions in TNBC and TPBC, where notable differences in fragmentation and methylation patterns were observed when compared with controls and other subtypes. Among the fragmentomic features evaluated, Fragment Size Ratio showed noticeable difference for TPBC and TNBC but did not demonstrate the same level of distinction for LB or HER2+. In methylation, end motif and breakpoint motif analyses, observable patterns were also more prominent in TNBC and TPBC, while clear differentiation was not consistently evident for LB and HER2+ within this dataset. The integrated analysis of cfDNA hypermethylation and fragmentomic features reveal distinct fragmentation signatures in TPBC and TNBC, emphasizing the potential of cfDNA-based assays for precise, non-invasive molecular subtyping and the development of liquid biopsy strategies for early detection and disease monitoring.